453 - Impact of Patient-Level Factors on Neurodevelopmental Follow-up Attendance Among High-Risk Newborns in Washington State

Publication Number: 4444.453

Sara K. Neches, University of Washington School of Medicine, Seattle, WA, United States; Olivia C. Brandon, University of Washington School of Medicine, Seattle, WA, United States; Kanika Nallaseth, University of Washington, Seattle, WA, United States; Tommy J. Yang, University of Washington School of Medicine, Sammamish, WA, United States; John Feltner, University of Washington School of Medicine, Seattle, WA, United States; Krystle Perez, University of Washington School of Medicine, Snohomish, WA, United States; Kendell German, University of Washington School of Medicine, Seattle, WA, United States; Ulrike Mietzsch, University of Washington School of Medicine/ Seattle Children's Hospital, Seattle, WA, United States; Niranjana Natarajan, University of Washington School of Medicine, Seattle, WA, United States; Cindy I. Trevino, Seattle Children's, Seattle, WA, United States; Sandra E. Juul, University of Washington, Seattle, WA, United States; Brent Collett, University of Washington School of Medicine, Seattle, WA, United States; Gregory C. Valentine, University of Washington School of Medicine, Tacoma, WA, United States

Background: Neurodevelopmental follow-up (NDFU) is critical for small and/or sick newborns globally. Yet, nearly 50% of high-risk infants referred for NDFU do not attend, exacerbating underlying disparities in care and outcomes. Understanding patient-level predictors of NDFU may help inform retention and outreach efforts for vulnerable populations.

Objective: To evaluate factors predicting NDFU for infants discharged from an academic neonatal intensive care unit (NICU).

Design/Methods: A retrospective chart review was performed for infants meeting criteria for high-risk NDFU who were discharged from a high-risk delivery center between 01/01/2022 and 07/31/2025. Infants were excluded if they were transferred out and discharged from another facility. Sociodemographic variables, clinical variables, the Centers for Disease Control and Prevention's Social Vulnerability Index (SVI), and attendance/nonattendance to NDFU were collected (Table 1). Clinical Risk Index for Babies [CRIB II] score) was used as a marker of illness severity. Variables were analyzed via descriptive statistics, forward-step Akaike Information Criterion (AIC) regression and graphical network analysis to identify significant associations with NDFU within the first 12 months corrected age (CA).

Results: Among n=183 infants, 110 (60%) attended any NDFU before 12 months CA. Infants who attended NDFU had lower median gestational age at birth (29 weeks [27, 30 Interquartile Range]) and higher median CRIB-II scores (7 [4, 9]) than those who did not attend NDFU. All patients with hypoxic ischemic encephalopathy attended NDFU (Table 1). Therefore, subsequent graphical network and AIC analyses focused specifically on preterm neonates. Graphical network analysis revealed increased association of NDFU with higher CRIB-II scores (greater acuity) and lower SVI (reduced social vulnerability). Forward step AIC confirmed these findings, showing that NDFU was significantly associated with CRIB-II scores (aOR 1.17, 95% CI 1.06-1.31, p=0.004) and having private insurance (aOR 2.23, 95% CI 1.06, 4.80). Conversely, higher SVI scores were associated with lower odds of attending NDFU (aOR 0.84, 95% CI 0.73-0.96, p=0.01). Other variables evaluated were not significantly associated with NDFU.

Conclusion(s): Attendance at NDFU was lower among families with markers of social vulnerability and higher among preterm neonates with greater severity of illness after birth. Considering social vulnerability and tailoring NDFU to families' needs may increase participation in follow-up for those at greatest risk of loss to follow-up.

Table 1. Patient characteristics and attendance with neonatal follow-up*
Table_1.NDV.pdf*Criteria for neonatal follow-up: <32 weeks' gestation at birth and/or <1500g, hypoxic ischemic encephalopathy (mild, moderate or severe), confirmed sepsis or meningitis, cytomegalovirus, neonatal stroke, Congenital Diaphragmatic Hernia (CDH), history of Extra-Corporeal Membrane Oxygenation (ECMO). Note that CDH and ECMO patients are not represented in this data set.
Continuous variables shown as median (interquartile range) were compared between groups using the Mann-Whitney U-test.
Categorical variables shown as n (%) were analyzed using Fisher's exact test. p<0.05 is considered significant, and bolded p-values represent those <0.05 to visually demonstrate significance.
1) NDFU = Neurodevelopmental Follow-up
2) Available for analysis of maternal smoking, substances, or alcohol 72/73 in the no NDFU group and 107/110 in the any NDFU before 12 months group
3) Available for analysis of prenatal steroids 67/73 in the no NDFU group and 98/110 in the any NDFU before 12 months group
4) Available for analysis of patient ethnicity 62/73 in the no NDFU group and 107/110 in the any NDFU before 12 months group
5) Available for analysis of intubation/chest compressions and vasopressor use 109/110 in the any NDFU before 12 months group
6) Bronchopulmonary Dysplasia as defined by the Jensen criteria 2019 (Jensen et al 2019): Grade 1 (mild BPD) requiring nasal cannula with a low flow rate of ≤2 liters per minute, Grade 2 (moderate BPD) requiring nasal cannula with a high flow rate of > 2 liters per minute, nasal continuous positive airway pressure (CPAP), or non-invasive positive airway pressure (NIPPV), or Grade 3 (severe BPD) requiring invasive mechanical ventilation.


Figure 1. Graphical network showing factors associated with any follow up before 12 months corrected age (CA).
Lines indicate significant associations between parameters from negative (red) to positive (blue). Thickness of the line indicates the magnitude of the association. Each variable is adjusted for the others in the graphical network model. Notably, only decreasing SVI and increasing CRIB-II score were significantly associated with NDFU within the first 12 months CA. N=145 infants were analyzed within the graphical network.

Figure 2. The adjusted ORs for predicting any NDFU before 12 months corrected age (CA) with variable selection following forward step AIC regression modeling.
Of note, SVI, CRIB-II score, Hispanic ethnicity, and private insurance were selected by the forward step AIC regression model which initially included all variables within the graphical network analysis (Figure 1). For continuous variables, it is the aOR per increase in 1 unit. For every 1-point increase in CRIB-II score, there was 17% increased odds of NDFU before 12 months CA. Of note, SVI was multiplied by 10 in the regression model; therefore, for every 0.1-point increase in SVI, there was 16% reduced odds of attending NDFU within the first 12 months CA. NDFU: neurodevelopmental follow-up. aOR: adjusted odds ratio. CI: confidence interval. SVI: Social Vulnerability Index (SVI). Purple shows significant associations and orange is not significant. aOR: adjusted odds ratio. CI: confidence interval. SVI: Social Vulnerability Index (SVI).